Serological response in RT-PCR confirmed H1N1-2009 influenza a by hemagglutination inhibition and virus neutralization assays: an observational study

Abstract

Background: We describe the serological response following H1N1-2009 influenza A infections confirmed by reverse-transcriptase polymerase chain reaction (RT-PCR).

Methodology and principal findings: The study included patients admitted to hospital, subjects of a seroepidemiologic cohort study, and participants identified from outbreak studies in Singapore. Baseline (first available blood sample) and follow-up blood samples were analyzed for antibody titers to H1N1-2009 and recently circulating seasonal influenza A virus strains by hemagglutination inhibition (HI) and virus micro-neutralization (VM) assays. 267 samples from 118 cases of H1N1-2009 were analyzed. Geometric mean titers by HI peaked at 123 (95% confidence interval, CI 43-356) between days 30 to 39. The chance of observing seroconversion (four-fold or greater increase of antibodies) was maximized when restricting analysis to 45 participants with baseline sera collected within 5 days of onset and follow-up sera collected 15 or more days after onset; for these participants, 82% and 89% seroconverted to A/California/7/2009 H1N1 by HI and VM respectively. A four-fold or greater increase in cross-reactive antibody titers to seasonal A/Brisbane/59/2007 H1N1, A/Brisbane/10/2007 H3N2 and A/Wisconsin/15/2009 H3N2 occurred in 20%, 18% and 16% of participants respectively.

Conclusions and significance: Appropriately timed paired serology detects 80-90% RT-PCR confirmed H1N1-2009; Antibodies from infection with H1N1-2009 cross-reacted with seasonal influenza viruses.

Figure 1. Hemagglutination inhibition titers by time from illness onset to sample collection.

Includes 267 samples from 118 individuals. Samples taken before onset in sero-incidence cohort participants are grouped (as <0* days), as are those taken after 70 days (≥70 days), with other observations summarized in 5 day intervals up to 29 days, 10 day intervals from 30 to 69 days. The number of samples in each interval is in brackets; intervals marked with # include 7 samples from the 3 individuals whose date of positive PCR test was used instead of onset dates: <0 (2 samples), 20 to 24 (1 sample), 40 to 49 (2 samples) and ≥70 (2 samples). Colored stacked bars give the proportion with titers of 10, 20, 40 and ≥80 while the line denotes the geometric mean titer with error bars depicting 95% confidence intervals. The upper limit is off the scale for days 10–14 (518) and days 30–39 (356).

Figure 2. Seroconversion on hemagglutination inhibition assay, by timing of sample collection.

Titers between baseline and follow-up (FU) sample is compared, for all participants regardless of time from onset to follow-up sample (n = 90), and restricting to participants whose follow-up sample was taken 15 or more days after onset (n = 68), with error bars depicting 95% confidence intervals; columns of different colors depict different cut-off points for time from onset to baseline sample. Seroconversion was defined here as a four-fold or greater increase in antibody titer.

Figure 3. Comparison between hemagglutination inhibition (HI) and virus neutralization (VM).

A) baseline sample titers (n = 45 for HI, n = 45 for VM); B) follow-up sample titers (n = 45 for HI, n = 44 for VM); C) fold rise in titer for follow-up relative to baseline titer (n = 45 for HI, n = 44 for VM). In all three panels, lines denote the reverse cumulative distribution with error bars representing the 95% confidence intervals.